Mixed feelings about broadband caps

[Expanding on some thoughts I've made in comments elsejournal recently.]

Comcast has just made official what they've been hinting at for quite a while: they will be capping residential monthly bandwidth at 250GB. Needless to say, this is causing some consternation in the blogosphere.

On the positive side, they're actually giving a concrete number, as opposed to being vague about it, which is a real improvement -- there have long been reports of them tapping heavy users on the shoulder and saying, "Too much". If that's going to happen, I'd much rather have them saying publically what the limit is rather than applying it secretly and arbitrarily. And in their defense, 250GB *is* a lot of traffic: I'd be pretty hard-pressed to use that much with current technologies.

More importantly, the economist in me doesn't mind the idea of bandwidth caps. The notion of "all you can eat" is bad economics: by failing to set a price on consumption, it encourages people to be stupidly wasteful, and that will usually lead to bad results. By saying explicitly "this much bandwidth will cost you this much", it allows people to see what their consumption is costing, and to compare plans appropriately. In principle, competition should drive the cost-per-bit down to appropriate levels.

(And perhaps even more importantly, it's a move towards treating Bits As Bits, which is a healthier way to think about the Internet than trying to distinguish one kind of traffic from another, as they've been doing lately.)

OTOH, in practice there's a big snag here -- this isn't a particularly competitive market. It's not *too* bad in the Boston area: we've got three big players (Comcast, RCN and FiOS) duking it out in many towns, which is enough to keep them somewhat honest. But a lot of places have only two, which is dangerously cozy, and many only have one, which typically leads to things being more expensive than they should be. Even with three players, frankly, the market is smaller than I like, especially when those players have shown a tendency towards acting as a block when convenient.

So the situation does make me a bit nervous: in the face of weak competition, Comcast could potentially abuse their pricing power. The only reason I'm not *really* worried is the details of this particular market -- FiOS is the hungry new entrant into this market, and one of their best competitive advantages is bandwidth. (FiOS' network is in some respects much better than Comcast's.) I expect this to keep Comcast honest, lest they hand Verizon a big competitive advantage. But it's worth watching developments in this space closely, and it makes it even more important to fight against the industry's constant attempts to stifle new competition...

Mixed feelings about broadband caps

[Expanding on some thoughts I've made in comments elsejournal recently.]

Comcast has just made official what they've been hinting at for quite a while: they will be capping residential monthly bandwidth at 250GB. Needless to say, this is causing some consternation in the blogosphere.

On the positive side, they're actually giving a concrete number, as opposed to being vague about it, which is a real improvement -- there have long been reports of them tapping heavy users on the shoulder and saying, "Too much". If that's going to happen, I'd much rather have them saying publically what the limit is rather than applying it secretly and arbitrarily. And in their defense, 250GB *is* a lot of traffic: I'd be pretty hard-pressed to use that much with current technologies.

More importantly, the economist in me doesn't mind the idea of bandwidth caps. The notion of "all you can eat" is bad economics: by failing to set a price on consumption, it encourages people to be stupidly wasteful, and that will usually lead to bad results. By saying explicitly "this much bandwidth will cost you this much", it allows people to see what their consumption is costing, and to compare plans appropriately. In principle, competition should drive the cost-per-bit down to appropriate levels.

(And perhaps even more importantly, it's a move towards treating Bits As Bits, which is a healthier way to think about the Internet than trying to distinguish one kind of traffic from another, as they've been doing lately.)

OTOH, in practice there's a big snag here -- this isn't a particularly competitive market. It's not *too* bad in the Boston area: we've got three big players (Comcast, RCN and FiOS) duking it out in many towns, which is enough to keep them somewhat honest. But a lot of places have only two, which is dangerously cozy, and many only have one, which typically leads to things being more expensive than they should be. Even with three players, frankly, the market is smaller than I like, especially when those players have shown a tendency towards acting as a block when convenient.

So the situation does make me a bit nervous: in the face of weak competition, Comcast could potentially abuse their pricing power. The only reason I'm not *really* worried is the details of this particular market -- FiOS is the hungry new entrant into this market, and one of their best competitive advantages is bandwidth. (FiOS' network is in some respects much better than Comcast's.) I expect this to keep Comcast honest, lest they hand Verizon a big competitive advantage. But it's worth watching developments in this space closely, and it makes it even more important to fight against the industry's constant attempts to stifle new competition...